The overarching goal of this project is to identify and characterize genetic determinants of heroin abuse in large samples of African Americans and Caucasians by conducting (1) a case/control genome-wide association study (GWAS) of heroin abuse;(2) cross-population contrast mapping to help identify causal variants;and (3) replication analyses in independent samples. To achieve this goal we propose to capitalize on our current GWAS of HIV-1 infection among injection drug users (IDUs) (DA026141), and match the Urban Health Study's (UHS) 3,878 African American and 2,685 Caucasian heroin abuse cases to controls from publicly available data. The R21 phase focuses on identifying, obtaining, and matching optimal control subjects to cases. The R33 phase focuses on the genetic analyses of the derived case/control data. GWAS have had a number of replicable successes identifying genetic variants that contribute to the risk for complex diseases including for substance abuse (e.g., CHRNA5 with nicotine dependence). There are few GWAS of substance abuse, particularly for rarer high-risk substance abuse like heroin abuse. Although about 50% of the risk for heroin abuse is attributable to genetic factors and a number of molecular genetic studies have yielded intriguing results, no genetic variants have strongly replicated evidence of contributing to this genetic risk. To address this need, we will pursue the following R21 and R33 aims: Aim 1: To identify, evaluate, and acquire control samples to match to UHS heroin abuse cases for a GWAS. Aim 2: To develop analytic datasets from UHS heroin abuse cases and identified control samples. Aim 3: To evaluate genetic associations for heroin abuse in UHS cases (n=6,563) and derived controls among African Americans and Caucasians. Aim 4: To conduct cross-population contrast mapping of top GWA findings to select single-nucleotide polymorphisms (SNPs) for follow-up. Aim 5: To replicate primary findings in independent cohorts. Matching the 6,563 heroin abuse cases to repository controls, the proposed GWAS will be many times larger than any genetic study of heroin abuse and one of the largest of any drug abuse phenotype, allowing for detection of expected small to modest genetic effects. Thus, this GWAS is likely to discover, and to refine understanding of, genetic variants associated with heroin abuse, provide important clues to the biology of addiction, and indicate targets for further study and development of pharmacogenetic treatments. PUBLIC HEALTH RELEVANCE: Over two million Americans abuse heroin, at great personal and societal cost. About 50% of the risk for heroin abuse is attributable to genetic factors. This study will identify genes associated with heroin abuse among Caucasian and African Americans and will contribute significantly to addressing minority under-representation in genetic research. The results of this study may identify important biological pathways for addiction and targets for developing new treatments.